"How AI is Changing the Game for Companies of All Sizes" IBM's VP Technology + Director Startups

Global Vice President for Software and Technology partnerships at IBM. Hello everybody. Welcome to Saster. We're a little light probably due to the party last night, but hope everyone's having a great time and I'll just quickly introduce myself and hand it over to Raj. So, Kylie Rutherford, I lead our startup ecosystem with IBM and we're incredibly excited to be at Saster for the first time ever with IBM to help you startups really harness the power of AI and transform in this era of, of data and AI. So I'll let Raj give his intro.

Thank you, Kylie. Good morning everyone. Hopefully everyone's enjoying the conference here. This is my first time attending and it's been pretty amazing thus far. So my name is Raj Datta. I am responsible for all software and technology partnerships for IBM. And today's topic, as Kylie mentioned, is going to be with regards to AI. And so, you know, I'm sure this has been the buzz, you know, throughout the conference and as we've seen over the past five years, five, six years, the growth of AI has been exponential, right? And as you can see on this chart, you know, it says that the growth of AI has happened more than 2 1/2x over the past five years. I would actually say in the past six months we've probably seen this number double as well, right? And why is it doubling? Right. This is a, we're starting to use AI as an area of getting the compet beating the competition, right? Each of us who have your own companies and things like that, it's a very competitive landscape and how are you going to get the edge?

Right?

And as I'm going out there and speaking to many, many customers, right, what we're seeing is every CXO out there is saying yes, we're going to adopt AI, we're going to, we need to move forward. This is critical for our business. But the problem that occurs is as you see by the stat, 79% are running to this issue. This issue is very simple, right? You look at it, companies will struggle to ensure that AI models are responsible, secure and free of bias. So you know, just a little bit of background for me, I actually just returned back to IBM about a month and a half ago and prior to this I, I was actually, I had my own software, CyberSecurity software company, Oak9 and this was one area that we were really struggling with is how do you have responsible AI, especially as a cybersecurity software company. And as we look at this, right, it's really about where is the data coming from? How are you training your models that you get responsible AI? And this was actually one of the reasons why I came back to IBM, because the focus of AI has been with us for 20 plus years. I mean, I'm sure most of you saw a long time ago when we, you know, Watson went on Jeopardy. And beat the Jeopardy. Champion, right? But it's evolved now. Now we have an enterprise grade software, right? And so as I'm talking to earlier stage companies, what I've also seen is, you know, companies are running into biases and issues with their modeling. I'll give everyone a very simple example and I'm not trying to throw shade, but you know, I'm sure most of you have, you know, chatgpt on your phones or have used it, right? If you take a look at it and you ask a simple question, you know, spell lollipop backwards. Take a look at that after this and I would love to get your response because it's not able to understand that. And why is it not able to understand that? It's because it was how the data was trained and the models that were built, right? So as we go through this, you know, we're looking at how organizations are wanting to embrace AI but feel that they don't have the right skill sets to get there. So Here you see 34% of the businesses lack the skill sets and knowledge for AI. I would say actually this was a stat from 2022. It's grown even more. And one thing that we've done here at IBM to prove it is we eat our own dog food. So over the past month, what we did across the organization, so this is 200,000 plus employees, actually did a WatsonX challenge. And what we did there is we had salespeople, we had technical people, we had marketing, we had our secretaries, we had everyone participate in this WatsonX challenge. And what they did was they created their own AI applications using WatsonX. So that shows you that where IBM is focused is we are enterprise grade. But we want to give the ability for everyone to be able to use it. Regardless of your company size, for anywhere from enterprises to startups, WatsonX is available for you. So that said, as we look at it, three simple areas that we focus in on is how do you create a competitive edge? Every software company is trying to do that today. That's one of the most critical areas we look at then is how do you scale AI across your business? Do you have an enterprise grade platform that allows you to do so? And then how do you advance trustworthy AI So this is your compliance area, risk security, making sure that the AI that you're putting out there with your software is responsible. And so that said, I'm gonna hand it off to Kylie, who's gonna go through some of pretty interesting use cases that we've come across that are just, you know, repeatable and usable across many different industries.

So awesome. Thanks, Raj.

So.

So before I jump into use cases, I'm going to add on to a piece of information that Raj was sharing earlier about how IBM is really helping companies change right now with, with, with our AI. So to his point, you know, Raj was at IBM, he went to a startup, he came back. I also was previously with Amazon Web Services for a really, really long time. And what brought me over to IBM was truly this massive opportunity in the market with. And I think what's really important and really exciting is we're doing things in the market like coming to Saster. So this is, you know, we're flexing a really important muscle that says, how can we really be helping founders and co founders and CEOs and, you know, folks, some of the partners that we have in the crowd today really get their business up and running in a way that delivers immediate value. And I think in broader strokes, like, there can be, you know, some resistance to AI or, oh, maybe it's going to be a little scary. But what comes along with that is that change, change is inevitable. And so what's really, really important as we go through these use cases is that you think about what are the necessary changes you need to make in order to grow. Right. Companies and people, we often don't grow without change. And so thinking about when we're looking at these use cases, how can it apply to your business? And we definitely want to have those conversations with you, you know, after our session. So definitely come find us, come chat with us. We are really, really excited to speak with each and every one of you. So do you want to add anything to that?

I was just going to say, like, well, if you're not doing it today, I can guarantee your competitor is. Yeah, it's a race to see who adopts it the fastest and does it in the best manner. Right?

Absolutely, absolutely. And, you know, part of being in that sort of landscape is you're already up and coming to become relevant. Maybe you already are. And so you, you're gonna be looking at things a little bit differently. So rather than how do I transform maybe legacy systems, you might be just saying, how do I build in a way that's built for tomorrow, how do I build in a way that's built for enterprise? So when I land multiple enterprise clients or I'm having those conversations, we have everything in place to scale up very, very rapidly. So let's talk about how you can do that because there are a couple really, really key use cases that we're seeing and we're directly supporting partners and companies that are doing this today. So the first one I'll chat a little bit about is talent. So none of us can build a really successful company if we're not thinking about our talent. But acquiring the right talent is a difficult task. It takes time, it takes energy. And you also have to be really cognizant of how are you reducing bias in hiring talent? How are you hiring diverse teams? Teams? What are the steps that you're taking that's going to build your rock star a team? Right. So when we're thinking about that with AI, there's a level of productivity that can be improved when you're leveraging AI. So things like, you know, content generation, I can be building your job posts for you. It can be proofreading and editing and looking for certain words or verbiage that could potentially create a negative perception from your audience. Right. You want to encourage all types of people to apply for a particular role if they meet the qualifications. So these are just the starting points of how AI can really start to infuse and help your business when it comes to talent.

Well, I was going to say I must have tricked IBM's AI models because they ended up hiring me back. But you know, one interesting thing also with regards to the HR modeling is even at IBM, as I mentioned, you know, we eat our own dog food and we actually ours is essentially an HR chatbot that majority of our employees use for HR related questions. And we've trained it over many, many years where that now is the de facto standard for all HR related items.

And this is a really, really great point because it segues into customer service very, very well. Our chatbot, for example, can resolve 70% of the employee cases that would typically be required to be answered by a person. You know, a lot of the questions that we get, whether it's talent or customer service, a lot of those are repeatable. So this is really important as well that when we're looking at the customer service landscape, there's your customer profile, who is your customer, how many? Like there's almost always a case management, a support desk, a help desk of some sort. You have agent management. Right. And when you're looking at the cost of that, whether you're outsourcing or you're doing it in house and it's homegrown, these are things that, that you can evolve, you can modernize. You don't necessarily need to spend those costs on agents. So rather than do that, you don't look at things like retrieval, augmented generation that takes data across processes and things like FAQs and answers that with generative AI and conversational AI. So love how we're doing that at IBM. That's actually our ask. HR is something internally that I found very impressive when I came back. And ultimately you can still get a hold of a person if you really, really need to, right? And so that's the balance is you need to understand, well, what are the pieces based on my customers, based on my use case that are really, really important to be using AI and then where can I extend and spend more valuable resources? What are those activities they should really be focused on? And then the third one is application modernization. I think for startups I would position this slightly differently and say you're already building for the modern environment, right? So there might be some of you who maybe you started one company and then you made a major pivot and you're like, actually no, I have some really good components, but I'm going to take a slightly different direction. And you had to modernize. So things like automating code generation, using your developer organization more wisely is critical here. So things like using code assist to generate and automate the design of code, or proofreading your code, or transforming your code from a legacy system to a.

New structure, suddenly everyone's becoming a developer, right?

Yeah, yeah, everyone's a developer. Everyone does AI. Like these are themes that we're seeing just across the booths and, and here at Sasser today. So let me go ahead to the next slide here. I want to highlight two companies that we're working with who have done an amazing job of this and really the goal is to bring to life how they did it, right? We can talk about the use cases, but understanding what they went through and what was really important to them. So first I'm going to highlight Alice Chrissy, the co founder and CEO who's up in the front. So thank you so much for coming, Alice. Alice is our co founder and CEO of Ova Medical and Ova Medical is a med tech startup company that is focused on a really, really important and sensitive topic which is solving questions around fertility and health. So as you can imagine, when you're dealing with fertility, this is a sensitive issue. The people who are coming to Ovum looking for answers have probably been trying to get pregnant for a long time, and they're trying to find answers. And so they're in a vulnerable place. And so when they come to a place like Ovum Medical, they're looking for not only answers to their medical questions, but they need it to read with empathy. And so that was one of the first things I remember Alice told me on the phone, she's like, I don't care what the technology does. If it's not empathetic, it's not going to work. And we said, this is beautiful. We want to work with you, because sentiment is something that can absolutely be trained in our models, and we can highlight the sentiments that are most important to you. So not only did we leverage our tools and systems to build a solution that would come in and help her scale 60,000 patients, both their Q and A, with medical professionals, as well as actually converting and booking appointments. So you can just imagine all of these people, we resolve some of these cases with AI while also keeping a really, really empathetic tone, which I think is absolutely critical. So that's a little bit about Ovum. And let's chat more about one more company I have here, Make Music Count. So this is a. I love Make Music Count. Really, really interesting technology product. They're focused on using, taking math and making it fun for students. So Marcus, the CEO and co founder of Make Music Count, he saw a challenge in the school system where kids were starting to really, I'm trying to think of the right word. They didn't want to do their math.

Right.

They didn't like math homework. It was hard. It was not fun. And he said, we need to think through a new way to educate students on a really, really critical topic. And so he infused music and pianos and all these things to basically say, hey, let's help kids learn math through music. So really, really exciting platform. He has again, upward of 60,000 students that signed up in a very, very short time. And so Marcus, of course, is like, how do I connect the dots between all the students, all the facility leaders at the schools? And then also I'm trying to onboard more schools to leverage Make Music Count. So we're working with them right now on how we can embed AI into their products so that they can scale faster and go across schools, which is really, really exciting. So with both of those examples and only a few moments left of your time, I want to hand it over to Raj to share a really, really special opportunity that we have built particularly to help you design your AI strategy.

Yeah. And so, like, as we continue to grow in the AI market, I want the opportunity to offer to everyone here the ability to try out our Watson X and give it a shot. Because. And either you'll love it or you won't. And if you don't, I would love to hear from you personally.

Right.

But my gut tells me that you'll really find value in this. And so, you know, as Kylie mentioned, this is a great opportunity for each of you to try it. And you know, like, as you've seen, like, there's really interesting announcements that we've been making over the past few months. Right. With our partnership with Meta and Llama 2 models. We recently also announced a co investment with Salesforce Ventures. With hugging face.

Right.

We are attacking the space in a programmatic way.

I'm sorry.

And Llama two.

Llama two.

Yeah.

We're attacking this space in a very programmatic way because we want to do what's right for companies that are out there. And I would welcome each of you to give WatsonX a try and then we would love to hear more and we'd love to have you up on stage with us next year at Saster talking about how you implemented WatsonX and how your AI strategy really grew over the past year. So that said, I just want to thank everyone for spending some time with Kylie and I and we'd love to chat with you afterwards.

Awesome. Thanks everybody. Please welcome to the stage michael grinich.

CEO at workos, and krithika muthakumar, head.

Of marketing at OpenAI.

Good morning, Saster. How many of you remember where you were in March of 2020? I know where I was. I was sitting at home, probably like most of you on my computer watching the news as the very beginning of the pandemic shutdown happened. When we were told it would only be two or three weeks wait for it to wash out. But I was waiting for something else. We were working on a new company that had been in stealth for about a year and a half at that point called Work os. And we were getting ready to launch it. And our launch plan for workos was not to go get super exclusive story from a top publication. It wasn't to pay a PR firm or do any sort of blowout. We launched Work OS on this niche website for developers called Hacker News. And our goal was to be on the top of Hacker News all day. This niche, niche audience with developers at the end of that day, we had no snarky comments from developers. And that was really the gem that we knew that we were onto an idea.

Now, it's kind of remarkable what Michael's saying, right? To launch a company with a single tweet or a post on Hacker News seems bizarre, but it's actually kind of par for the course when it comes to developer facing companies. If we were giving this talk even a couple years ago, we'd be trying to convince you all as SaaS companies to be adding a platform form or an API to your products. But that train has left the station. Most of you in the audience are working on developer tools, developer facing companies, or your product already has a platform or an API, but what's still a little bit of a black box is what are the actual tactics, right? What are the most effective ways to go to market with this very unique audience of developers or very technical buyers? And so that's what we're going to share with you today. As Michael said, work OS and his previous company, Nihilus, also launched on Hacker News as their tactic. But that's not dissimilar to how Retool launched, how Stripe launched. I'm also very new to OpenAI. I'm not going to talk as much about OpenAI today, but there's a lot of learnings that we've kind of hard won over the past decades working on developer facing products that we want to share with you all. We have about 20 minutes for this talk, so we're just going to catch the very tippy table TP top of the iceberg, but give you sort of a broad spectrum of different areas and topics that you might want to consider when it comes to going to market with developers. And we hope that whets your appetite for what this could unlock for you and your business. Every company is becoming a platform company, so I hope there's something for everyone to take away. Let's start with user acquisition.

All right, let's jump in. So acquiring users, the first step when you're launching a business is actually to get users right. Eventually they'll pay, but you want to get them to acquire them first. Acquiring users for developer products is a bit different than acquiring them for other SaaS products where you might run Google Ads or like I said, hire a PR firm. If you were launching a traditional company, maybe the dream would be to get, you know, a feature piece in the New York Times, get their commentary on your product, and that would really set the stage for the consumer sort of perception of what You've launched. In the world of developer tools, developers are not reading the New York Times. They're on the platform formerly known as Twitter now. X. That's where developers congregate. That's where they share ideas. That's where they talk about what's new and what's cool. So you really need to win them over on that platform. Similarly, if you were to launch a new company, you might try to get on the front page of the Wall Street Journal. Like a front page piece, you know, would be a traditional marketing win. Developers aren't reading the Wall Street Journal. The front page the developers care about is what I described, Hacker News, which is sort of like Reddit. It's a message board for developers and technology enthusiasts looking for the great new thing. And there are countless companies that have launched just with the show hn post. These are a couple ways that you can acquire those customers, acquire those users. I should say. I really want to emphasize that it's super important to be authentic with developers and that really means being technical with them. When you create content, you want it to be authoritative and really like high integrity. With developers, it's really hard to fake this. If you're not technical, you need to get an engineer actually to usually write it or review the content. You can also give conference talks, have an engineering blog. One of the most important things is for leadership to directly get involved. For those of you familiar with Stripe, Patrick Collison, the CEO, still regularly gets on Hacker News and just comments like the rest of us on new press releases about Stripe for new feature announcements. And developers love that. They love hearing directly from technical leadership early on. You want to avoid all the typical marketing tactics. No search engine marketing, no webinars. The golden rule here should really just be provide as much value as you can and ask for nothing in return. Might seem counterproductive, but it's a long game that you're building here and if you do this right, it should work. Here are a bunch of different brands and different companies that have invested in their developer experience and developer audience. Early, early on you probably notice a theme of the colors between the purplish blue and the dark mode. That's definitely in fashion. But the audience that they've built and the momentum that they've built, they've been able to turn into really large meaningful businesses down the line. So it's definitely worth investing in. Last thing I want to say about this developer brand has to be built first. You can't add this later. You can't slap it on at Some point in the future you can't build a great product and then kind of hope to win over developers later on. It's really about building trust with them. It takes a long time to create it and build it up and you can be fast to lose it. So definitely, I would say try to invest in it as early as you can and then continue to maintain it over time. It's almost like kind of winning a fashion show with the developer audience. There could be a whole talk about this, but we're going to move on. Secret number one, though, focus on the brand first with developers and sequence all your acquisition tactics after that. Now we'll talk about outbound.

Well, Michael and I have been friends for some time now, so I think he'll be fine if I disagree with him in that you want to sequence all of traditional marketing mostly after the brand building and the sort of goodwill with developers. But one thing that you have to sequence upfront is outbound sales. And the reason that I say that is that you might actually think this is quite surprising for developers because you think they hate outbound sales. They hate getting emails in their inbox. But they there's this sort of misnomer in the market that if you build it, they'll come, you build this great product, this great developer experience, and of course the developers are going to bang down your doors and try to use it and go evangelize it to their friends. Folks, that's a pipe dream. The way that you want to think about this is that you want to hone your message. Sales doesn't lie. When you do outbound sales and you send these cold emails to people who have never heard about your product, it actually validates your messaging, your positioning, the value that developers might get from your product. So here are a couple of lessons at outbounding to get it right. One is warm. Outbound is always better than cold outbound. So instead of just spraying and praying to a bunch of audiences that you find from Apollo or from Lead Genius, instead you put your investor networks to good use. You can of course use ChatGPT Enterprise, which we launched last week, to create very custom emails to the audiences that you're trying to get to send those off to your investors and make sure that they're forwarding them onto the right folks in their network. Don't rely on just a sales team or a BDR function in order to be doing this outbounding. A founder to founder email is much stronger or even like a CTO to cto, email is much stronger in Terms of the results you might get shape your position through that iteration. What sticks and what doesn't will be very apparent. And that sort of iteration cycle is very short when it comes to outbounding, so. So that's another way to pressure test your messaging with technical audiences. Find the right targets. Because your practitioners, your developers who are actually using your product may actually be quite different than the decision makers who might be engineering managers. CTOs or even CFOs are often in the mix these days when it comes to the go to market motion and then scale the process alongside as you start to find your positioning. Oftentimes in developer platforms you have a cold start problem. You have this blank canvas. You're giving developers a box of Legos and telling them they can build whatever they want. And sometimes it's helpful to provide blueprints for success. So as you start seeing patterns in how people are using your product. We did this at Retool. We created patterns for how sales teams were using Retool, how healthcare teams are using retools, how people in the transportation vertical were using Retool. And that was a great way to increase the effectiveness of outbound. So sales doesn't lie. PLG is sort of par for the course these days. It's de rigueur when it comes to developer marketing. But Outbound is a great way to pressure test it in the early days and as you're scaling a new product. The other thing that I want to talk about is pricing. So how many of you launched a new product or feature in the last year? Great. How many of you changed your pricing in the last year? Okay, fewer of you. So I'm going to go out on a limb and say your pricing is probably wrong. These days it seems like most companies are using usage based pricing models and that's a good, great way to think about developer facing products. But increasingly there's more hybrid models that are coming into play. People are using seed based pricing as well as usage based models and conception models. And that's really important to think about. But there's not one way to fix your pricing. What I'm actually going to say instead is that you really need to experiment. So here's an example from Vercel, when it was even back when it was called Zeit, they changed their pricing. They tried a bunch of different, different pricing models over the years and that iteration has gotten them to a place that is much more solid for what they want to create. Pricing is always a two way door, right? It's a two way Handshake. On the one hand you're trying to communicate the value of your product and on the other hand your audiences are returning the favor and telling you where they find value in your product. So that handshake is really important and the only way that you'll get there is with experimentation. Retool has a free model. When we first started a two week free trial and we experimented with introducing a free plan that was free forever and that actually increased our top of funnel quite a bit. We lowered our pricing later on and we also introduced a new self serve business model where you did not have to talk to a sales team member in order to get on those more advanced plans. That also increased both our top of funnel and closed one rates for our sales teams. So again, the reality here is that if you are changing your product and you're not changing our pricing alongside, they're probably not in lockstep and you should.

Revisit pricing, pricing, pricing. Another whole topic. All right, number four, let's talk about product portfolios in the cross sell. So if any of you have talked to VCs or investors about product strategy, oftentimes they'll say you want to just do one thing and do it really well. Focus, focus, focus. And I think the only thing better than one product that does really well actually is a bunch of products that do well inside of a company. So I want to talk about the value of having a multi product portfolio, actually how you can execute on it. And before I jump into this, just want to share with developer platforms. This is possible with a lot of other companies. You actually might not be able to take your products and break it into pieces, but it's actually possible to do this as a component system with developer platforms. This is unique to that type of, of a strategy. So why pursue multiple products, more products, larger tam, you got more stuff that you can sell, you can address a wider audience. You also have more entry points to sell a customer. Maybe they don't need product A, but they need product B. You can start with product B. Eventually they might buy product A. Right? So you'll have that expansion effect. With lower energy. It's always easier to sell to an existing customer that you have and as you get, you know, accumulate multiple products that you're selling, you really, really have this distribution advantage. You already have a customer base that you can go sell to compared to single products. If you look at companies like Microsoft, over time they've executed on this really, really well. So there's obviously a big advantage to this. What Companies have done this well. Stripe has done this really well. Stripe started with payments, right? Just the credit card charge. They added invoicing, they added payouts. There's a bunch of different Stripe products now. Even with work os, we're four or five years old, we already have multiple products. We started with authentication, we added user management, we have a product run, audit, logging, all needed by the same customer base. Atlassian has done this too. You might not love their products like Jira and Confluence, but they've been really successful at it, right? And built many, many different ones. And perhaps the most successful developer platform multi product company is AWS. They started with storage. A lot of people forget S3 was their first product. Then they launched EC2, then network stuff and Route 53 and Yada Yada. And today Amazon has a lot of products, right? They've executed on this extremely, extremely well. They pretty much have everything you need to build a modern SaaS product. So it's really effective. You can execute on it. What are the pitfalls to watch out for? Well, initiating this too early actually might split your focus away from your core first business. If you think about kind of a rocket taking off, the first stage has got to burn out before you can get to the second stage. So you want to do it early, but not too early. Secondly, your engineering team might just want to work on the new stuff, new shiny stuff. So you have to figure out how to split your bets with the older product that you're supporting. Sometimes it's just generally harder to justify new bets when you have a business that's already succeeding. This is the classic innovators dilemma. If you've ever read that book, it's hard to build that conviction when you have revenue coming from existing products. And I think this one's actually the most scary pitfall is that if your two products or multiple products actually have a different set sales motion or a sales culture, you actually might not get the benefits that I mentioned. The previous slide around that kind of distribution advantage or sales advantage. So you need to make sure that the way that people adopt and the way that you sell it is similar. You know, Okta acquired Auth0 a few years ago, very different sales motions and they've had a really, really tough time integrating those products together. So huge upside if you can do it well. A lot of downsides if you don't do it well. But secret number four is it can really supercharge your product lineup and your revenue.

Okay, now I feel a little self conscious even talking about demand generation because I was an engineer before I was a marketer, so I'm quite a skeptical marketing leader. But demand gen, when you do it right, this is this misnomer that developers hate marketing and the reality is they just hate bad marketing. And so let's talk a little bit about those good examples. The neat thing is if you can sort of study the playbooks of what works and then sort of adapt that and really make it your own for that developer audience, it can actually be quite effective. So here's an example from mux. MUX created dmux, which is the industry conference for video professionals. This is a conference that did not exist before, but it's now the sort of preeminent gathering for that community to come together. A really effective way for them to build an audience and also even build a function and a role that they were then selling into field marketing and conferences just like the one that we're at, kind of get a bad rap sometimes because all you're trying to optimize for is the number of badges that you're scanning. Retool, when we started doing field events, we did not optimize for the number of badges scanned. We actually looked much more down funnel in terms of the pipeline that it was generating. That means we captured fewer leads, but we did deep custom demos to every single person who was coming by our booth because they're very technical audiences. All of the events that we went to were more than 2x ROI positive for us in terms of the deals that we won. Again, outbound emails, they get a bad rap. I talked about this before because most people pattern match to the bad stuff that you get in your inbox day in and day out. Instead, if you can write a really custom email that hits the nail on the head of a problem that an engineer is facing and how you're going to improve it, it can be really, really effective content. Again, there's so many listicles and bad crap out on the Internet. I really think that the sort of norm of what is actually on the Internet today is not great. But when it's again done well and it's done for, the thought of getting very technical content from an engineer to another engineer, especially in the developer space, is actually a broader lesson for all of us marketers in that we should be putting out content that speaks up to the audience, not down to them. Here's an example again from Retool, we put out a post about what's a CRM to engineers and it explained that very well to a very technical audience. And it did extremely well. Even today, eight years. Sorry. Five years after the post went out, it's still one of the top drivers of traffic to the company organically. I'm going to skip past a little bit of these examples because we don't have a ton of time left. But again, doing things that are altruistic for the community actually pay off a lot of dividends. They might be a little bit harder to measure in sort of a return window, but again, it's brand building as well as demand building. We ran an open source retreat at Stripe where we invited open source developers to come in and work on their projects from Stripe with a stipend that worked very well in terms of what they were able to accomplish in the time that they spent with us. And many other companies also picked up the baton and ran their own open source retreats. I love the 1.1.1 campaign from Cloudflare. Again, this is altruistic. They put it out there as a service to the community, not with the intent of driving demand, but definitely increase their brand and the number of leads that they got in the door. So learn those demand best practices and then deviate. If you do the thing that everyone else is doing, that's not going to serve you well with developers.

The last secret we want to talk about has to do with with. Org design, really how to think about building your initial sales team and support team together. So I would posit for developer products, most developers signing up just want to test something out, see the experience. There's a lot of free plans in developer products, kind of free to get started. Twilio pioneered this in a big way. What I would tell you is that early sales for developer products, it looks like great support. It doesn't look like sales, it looks like support. With usage based products, your long term value, really the terminal contract value you're going to sell to customers is actually dominated by expansion. That first month that developer puts in the credit card or the first deal you sign is going to be a very, very small amount of the long term value. The third year might be larger than year and 1 and 2 combined. So the whole game there is to keep that customer, keep them happy, keep them expanding, keep them building on your platform. You want to do this by optimizing for a really, really great customer experience via support. Later you can layer in traditional sales, you can always add it later. You can add it piecemeal, you can add it as needed, but you want to start with a great support and a great experience. This is a picture of Actually an early salesperson from Stripe, who we both know, Matt Yolowitz, who is actually here in front of a customer describing how Stripe works, just kind of laying out, here's the technical diagram of how all this kind of comes together. And this is a shared Slack channel between Stripe and Spotify, which was an early Stripe customer working on their integration, getting product feedback. Doesn't really look like typical sales, almost looks like a collaborator. We do this really heavily at work os. This is a screenshot of our Slack. We have hundreds and hundreds of shared Slack channels with customers. We're almost like an extension of their team that has turned into sales for us. Right? It started with support, started with this customer support motion, eventually drives revenue growth. You want to think about not always be closing, but always be helping. If you can help developers achieve their goals, the deal will happen. They'll build on top of you, they'll pay for it. And this begins with that early discovery relationship building, having a lot of empathy for that audience. Your initial sales team probably looks more like general purpose, individual contributors, people that are scrappy, willing to roll up their sleeves. Maybe more like a typical customer success team rather than a sales team. And if you do this really, really well, if you, if you give an extraordinary level of support, you can actually sell this. People are willing to pay for this because it's so unique and such a differentiator. So secret number six, great sales begins with great support.

So we're in our last 30 seconds of the time allotted to us. So again, we barely scratched the surface for what it means to really effectively go to market for developers. This is not meant to be a playbook. Right. The thing that we are trying to communicate is that what you have to do is diagnose what your company needs, what your developer community needs. So know your product, know your customer, and really connect the two in meaningful ways. So again, Michael and I really like to geek out about this stuff as when we talk. So if there's anything that you want to chat with us about, we're around for questions and then we'll be around the conference for the rest of the day.

You can catch probably both of us at the work OS booth right behind you. But after this. Yeah. Thanks so much.

Thank you. Sa.